sweden hairy

Zoological Journal of the Linnean Society (2001), 132: 441–468. With 8 figures
doi:10.1006/zjls.2000.0268, available online at http://www.idealibrary.com on
Phylogeny of Polygonia, Nymphalis and related
butterflies (Lepidoptera: Nymphalidae): a totalevidence analysis
SÖREN NYLIN1∗, KLAS NYBLOM1, FREDRIK RONQUIST2, NIKLAS JANZ1,
JOSEPH BELICEK3 and MARI KÄLLERSJÖ4
1
Department of Zoology, Stockholm University, S-106 91 Stockholm, Sweden;
Department of Systematic Zoology, Evolutionary Biology Centre, Uppsala University, Sweden;
3
15004-96 Avenue, Edmonton, Alberta, Canada;
4
Molecular Systematics Laboratory, Natural History Museum, Stockholm, Sweden
2
Received April 2000; accepted for publication July 2000
We investigated the phylogeny of butterflies in the tribe Nymphalini sensu Harvey 1991, comprising the genera
Vanessa, Cynthia, Bassaris, Aglais, Inachis, Nymphalis, Polygonia, Kaniska, Antanartia, Hypanartia, Symbrenthia,
Mynes and Araschnia. Evidence from the mitochondrial gene nd1, the nuclear gene ‘wingless’ and from morphology/
ecology/behaviour were used separately and combined to analyse relationships. Phylogenies based on the different
types of data agreed in many aspects of basic topology. We show that an analysis of only wing pattern characters
(based on Nijhout’s homology system) results in a topology broadly similar to the one resulting from analysis of the
complete matrix. We found support for a monophyletic Nymphalini, where Hypanartia may be the sister clade to
all other genera. Mynes, Symbrenthia and Araschnia together seem to form another basal clade. Evidence presented
gives only moderate support for a monophyletic Vanessa in the wide sense, including also Cynthia and Bassaris,
but strong support for the monophyly of the largely holarctic clade Aglais + Inachis + Nymphalis + Polygonia +
Kaniska + Roddia. Within the latter group there is strong support for a clade consisting of Aglais + Inachis and
for a second clade which includes Nymphalis, Polygonia (and its sister clade, the monotypic Kaniska) as well as
Roddia l-album (=Nymphalis vaualbum). As a consequence of this topology, Aglais is recognized as a taxon separate
from Nymphalis. We present a hypothesis of species relationships within the focal group of genera. We also analyse
and discuss the implications of excluding or including ecological data in phylogenetic tree construction, when the
 2001 The Linnean Society of London
tree is to be used for studies in phylogenetic ecology.
ADDITIONAL KEY WORDS: taxonomy – systematics – cladistics – mitochondrial DNA – nuclear DNA – phylogenetic
ecology – Nymphalini.
subset of the Nymphalinae, the tribe Nymphalini. The
remaining tribes in the Nymphalinae are generally
taken to be Melitaeini and Kallimini.
Nymphalini was suggested by Harvey (1991) to include the genera Vanessa Fabricius, Cynthia Fabricius,
Bassaris Hübner, Aglais Dalman, Inachis Hübner,
Nymphalis Kluk, Polygonia Hübner, Kaniska Moore,
Antanartia Rothschild & Jordan, Hypanartia Hübner,
Symbrenthia Hübner, Mynes Boisduval and Araschnia
Hübner. He also suggested that the most likely sister
group is either the Melitaeini (including e.g. Melitea
and Euphydryas) or Melitaeini + his Kallimini (including e.g., Kallima and Precis). In contrast, Ackery (1988)
included in his Nymphalinae the tribe Coloburini, as
well as a wider Nymphalini corresponding to Harvey’s
INTRODUCTION
The phylogeny of nymphaline butterflies is of considerable interest not only for taxonomists but also for
evolutionary biologists in general. The diversity among
these butterflies (e.g. wing patterns, behaviour and
host–plant associations) is intriguing, and the subfamily includes several important model groups for
ecological and evolutionary studies, such as Euphydryas, Precis and Polygonia. A step towards an
improved understanding of nymphaline phylogeny is
taken here by investigating the relationships within a
∗ Corresponding author. E-mail: [email protected]
0024–4082/01/080441+28 $35.00/0
441
 2001 The Linnean Society of London
442
S. NYLIN ET AL.
Nymphalini + Kallimini, but not the Melitaeini. Harvey (1991) placed the Coloburini in a different subfamily, the Limenitidinae, which he, however, did not
believe to be monophyletic. Hence, there is a controversy regarding both the monophyly of Harvey’s
Nymphalini and which are the closest relatives of this
group, if it is indeed a natural grouping of genera.
Within the Nymphalini, systematics has been in a
great deal of flux, although species circumscriptions
have remained fairly stable. The genera Cynthia, Bassaris, Aglais, Inachis and Kaniska have sometimes
been recognized and sometimes been included in other
genera. Cynthia (type species C. cardui (L.)) and Bassaris (type species B. itea (Fabricius)) are often included in Vanessa (type species V. atalanta (L.)),
although the revision of Field (1971) resurrected the
former two genera. The species of Aglais and Inachis
have sometimes been included in Nymphalis or Vanessa. For instance, Layberry, Hall & Lafontaine (1998)
recently argued in favour of including the two widely
recognized species of Aglais (urticae (L.) and milberti
(Godart)) in Nymphalis, because “Aglais might be a
highly modified (derived) group within Nymphalis”. In
contrast, Niculescu (1965) and others have argued that
Nymphalis in the strict sense (excluding Aglais) is in
fact more closely related to Polygonia than either of
these genera is to Aglais. The monotypic Kaniska
canace (L.) is sometimes treated as a Polygonia. We
will investigate which of these classifications is likely
to reflect phylogeny. We will also attempt to resolve the
controversy regarding the Holarctic species or speciesgroup Roddia l-album (Korshunov). This species is
most often referred to in the literature as either
Nymphalis vaualbum (e.g., Higgins, 1975) or Polygonia
l-album (e.g., Niculescu, 1965). As pointed out by Kocak
(1981), Nymphalis vau-album Denis & Schiffermueller, 1775 = vaualbum is a nomen nudum, and
the species name l-album of Esper (1780) is the correct
one. Recently this taxon has been placed in a new
monotypic genus; Roddia Korshunov, 1996 (Korshunov
& Gorbunov, 1995; Korshunov, 1996).
In this study we take a total-evidence approach
to resolving phylogenetic relationships. We combine
available data on morphology and ecology from literature with our own observations on all developmental stages, particularly adult wing patterns,
and with molecular data from both mitochondrial (nd1)
and nuclear (‘wingless’) genes. In a separate study we
analysed only the wing pattern data. This was done
because we have made extensive use of the recently
available system for suggesting homologies between
patterns in different taxa (Nijhout, 1991). We found it of
interest to investigate whether there is a phylogenetic
signal in this type of readily available data which is
congruent with other kinds of evidence, or whether
wing patterns are too affected by selection for, e.g.,
mimicry, creating homoplasy.
The phylogeny obtained in this study is intended to
be used in studies on the evolution of host plant
utilization in the Nymphalini. For this reason we also
studied more closely the effects of host plant data on
the resulting phylogenies.
MATERIAL AND METHODS
INGROUP
We strived to include all species in the focal group of
genera: Polygonia, Kaniska, Nymphalis, Aglais and
Inachis. This group is predominantly holarctic. The
number of species of Polygonia (in the narrow sense)
recognized varies, especially in North America, where
some relatively distinct populations are seen as either
species in their own right or as subspecies of P. faunus
(Edwards): hylas, smythi, silvius; P. gracilis (Grote
& Robinson): zephyrus; or P. progne (Cramer): oreas,
nigrozephyrus. The populations sampled by us represent all of the nearctic species recognized by Scott
(1986). In addition, there is probably at least one good
species south of the US, P. haroldii Dewitz, from which
we did not manage to obtain material. Of the four
palearctic species we included three: P. c-album (L.),
P. egea (Cramer) and P. c-aureum (L.). The fourth
species that was not studied, P. gigantea (Leech), is
found in China. The genus Kaniska holds a single
species, K. canace, which is often placed in Polygonia.
The same is true for Roddia l-album. Both of these
species were studied, l-album from both the Nearctic
and the Palearctic. However, as nd1 sequence difference between the two populations was found to be
negligible, and we saw no diagnostic differences in
morphology, we included only the Eurasian sample in
the final analyses. The same is true for the Holarctic
species Nymphalis antiopa (L.), where also samples
from both the Nearctic and Palearctic were initially
studied.
There are five clear species of Nymphalis, four
of which were studied: N. polychloros (L.), N. antiopa,
N. xanthomelas (Denis & Schiffermüller) and N.
californica (Boisduval). The fifth, N. cyanomelas
(Doubleday), is only very rarely found in the highlands
of Mexico, Guatemala and El Salvador (De la Maza E.
& White Lopez, 1986) and was not sampled for DNA.
The immature stages of this species are not known. It
was included in the morphological study on the basis
of a few adult traits, in order that we would be able
to suggest a probable phylogenetic position.
We studied the nearctic Aglais milberti (often placed
in Nymphalis) and the palearctic A. urticae. The other
species of Aglais described from Asia, A. kashmirensis
(Kollar) and A. ladakensis (Moore), in all probability
represent races of urticae, or at the very least closely
PHYLOGENY OF NYMPHALINI BUTTERFLIES
related species, such that their phylogenetic position
can be inferred from that of urticae. Inachis holds a
single species, I. io (L.), which was studied.
Within the Nymphalini, but outside of the focal
group, each genus was represented by one or more
species. Vanessa was represented by V. atalanta. Cynthia was represented by C. cardui and V. virginiensis
(Drury). Bassaris was represented by B. gonerilla (Fabricius), Antanartia by A. schaeneia (Trimen) and
Hypanartia by H. lethe (Fabricius), but we were
unsuccessful in extracting DNA from dry samples of
this species, so Hypanartia DNA in this study originated from H. lindigii (Felder). Symbrenthia was
represented by S. hypselis Godart (DNA, however, from
S. hypatia Wallace). Mynes was represented by M.
geoffroyi (Guérin-Méneville) and finally Araschnia by
A. levana (L.).
OUTGROUP EXEMPLARS
We included in the outgroup members of the tribe
Argynniti in the Heliconiinae, a subfamily well outside
of the Nymphalinae, in order to investigate monophyly
of the Nymphalini in the initial analyses. Argynnis
paphia (L.) of the Argynniti was used for the nuclear
gene ‘wingless’, morphology, ecology and behaviour,
but DNA had to be taken from Issoria lathonia (L.) in
the case of nd1.
As explained in the Introduction, the tribes Melitaeini, Kallimini and Coloburini are more likely to
be closely related to the Nymphalini, and in all the
main analyses we also included two species from the
Kallimini in the outgroup: Hypolimnas bolina (L.)
and Precis coenia (Hübner). Colobura dirce (L.) from
Coloburini was included in one analysis (see below).
SPECIMENS STUDIED
Morphology
The following taxa were followed throughout their life
cycle in the laboratory and the traits of immature
stages were investigated: Precis coenia, Araschnia levana, Vanessa atalanta, V. indica, Cynthia cardui, C.
virginiensis, Bassaris gonerilla, B. itea, Aglais urticae,
A. milberti, Inachis io, Nymphalis polychloros, N. antiopa, N. xanthomelas, Kaniska canace, Polygonia calbum, P. faunus, P. egea, P. c-aureum, P. satyrus, P.
gracilis zephyrus, P. interrogationis and P. comma.
Last-instar larvae, as well as pupae, of Mynes geoffroyi
preserved in alcohol were kindly provided by Darrel
Kemp. Traits of immature stages for the remaining
species included in the study, as well as some additional
traits of the entire set of species, were found in the
literature (references given below and in the character
list, Appendix 1).
443
External adult traits were also investigated in museum specimens of Argynnis paphia, Hypolimnas bolina, Nymphalis californica, Polygonia progne,
Antanartia schaeneia, Hypanartia lethe, Symbrenthia
hypselis and Mynes geoffroyi deposited in the Department of Zoology and Natural History Museum,
Stockholm. Mounted adults of N. cyanomelas were
studied at the Natural History Museum, London. Most
novel adult traits concern wing shape and especially
wing patterns, where we suggest possible homologies
based on the homology system presented by Nijhout
(1991). Several of these traits are depicted in Figures
6–8.
DNA
Butterflies for DNA analysis were collected by the
authors or obtained from colleagues (see Table 1 for
source populations). In most cases we used adult, pupal
or larval specimens which were killed by freezing and
stored at −70°C until analysis. In a few cases we
successfully used dry adult specimens from collections,
but most attempts did not yield a useful DNA extract.
Voucher specimens are stored at the Department of
Zoology, Stockholm, except for Symbrenthia hypatia
(wings retained by K. Fiedler) and Hypanartia lindigii
(extracted DNA kindly sent by A. V. Z. Brower). DNA
sequences will be uploaded to GenBank and are available from K. Nyblom upon request.
SOURCES OF CHARACTER INFORMATION (LITERATURE)
Data on internal adult morphology and ecology, as well
as several other traits, were taken from the literature
when we could not directly study them. Sources of data
for some particular traits are given in the character list,
but a few general sources can be mentioned here.
World butterflies: Harvey (1991); North America: Scott
(1986); South America: DeVries (1987); Europe: Niculescu (1965); Australia: Common & Waterhouse
(1972); Asia: Johnston & Johnston (1980); Nakanishi
(1988); Shirozu (1960); Shirozu & Hara (1960); Teshirogi (1990); Africa: Larsen (1991). In addition, some
sources were of special importance for specific taxa.
Mynes: Hawkswood (1990); Vanessa, Cynthia and Bassaris: Field (1971).
Concerning the male genitalia, we noted a controversy regarding terminology and homology of structures (see Niculescu (1985), between Niculescu (1965)
and Higgins (1975). Without substantially adding to
the scope of this study it was not possible for us to
study these traits independently. For this reason we
followed one author (Niculescu, 1965), reasoning that
homologies are more likely to be consistent within a
single author’s work, and Niculescu has the most
444
S. NYLIN ET AL.
Table 1. Origin of butterflies from which DNA was extracted
Ingroup taxa:
Polygonia c-album
Polygonia faunus
Polygonia progne progne
Polygonia gracilis zephyrus
Polygonia satyrus
Polygonia c-aureum
Polygonia interrogationis
Polygonia comma
Polygonia egea
Kaniska canace
Roddia l-album (‘Nymphalis vaualbum’)
Roddia l-album (‘Nymphalis vaualbum’)
Nymphalis antiopa
Nymphalis antiopa
Nymphalis polychloros
Nymphalis xanthomelas
Nymphalis californica
Aglais milberti
Aglais urticae
Inachis io
Araschnia levana
Antanartia schaeneia
Hypanartia lindigii
Vanessa atalanta
Bassaris gonerilla
Cynthia cardui
Cynthia virginiensis
Symbrenthia hypatia
Mynes geoffroyi
Inner outgroup taxa:
Precis coenia
Hypolimnas bolina
Outer outgroup taxa:
Colobura dirce
Issoria lathonia
Argynnis paphia
complete treatment of our study species. For the same
reason we have treated data as missing for non-European species except when homology was evident from
illustrations, as in the case of penis shape.
MOLECULAR ANALYSES
General
We initially intended to study only the mitochondrial
nd1 gene, which has been used in some earlier studies
of butterfly relationships (Aubert et al., 1996; Weller,
Pashley & Martin, 1996). A pilot study involving only
a few species also gave promising results; species of
the same genus ended up together in the phylogenetic
analyses and there seemed to be a reasonable
Stockholm, Sweden
Salmon River, Idaho, USA
Forest Co., Wisconsin, USA
Vale Mount., B.C., Canada
Blue Mountains, Washington, USA
Japan
Fayette Co., Tennessee, USA
Shelby Co., Tennessee, USA
Greece
Nagano, Japan
Manitoba, Canada
Ussuriysk Dist., Siberia, Russia
Stockholm, Sweden
Blue Mountains, Washington, USA
Öland, Sweden
Kirgisia, USSR
Jeff. Co., Colorado, USA
Blue Mountains, Washington, USA
Stockholm, Sweden
Stockholm, Sweden
Estonia
Cameroon
South America (DNA from A.V.Z. Brower)
Stockholm, Sweden
New Zealand
DeSoto Co., Mississippi, USA
Shelby Co., Tennessee, USA
W. Malaysia (wings retained by K. Fiedler)
Queensland, Australia
Shelby Co., Tennessee, USA
W. Malaysia
Costa Rica
Stockholm, Sweden
Stockholm, Sweden
resolution of relationships. However, as more species
were added much of this resolution was lost, and it now
seems that nd1 provides relatively little information on
relationships. This is despite the fact that we have
sequenced a considerably longer segment than in previous studies. For this reason we added the ‘wingless’
gene to the analysis. During the course of the nd1
work, this gene had emerged as a good candidate
for providing phylogenetic information at the level of
species and genera in butterflies (Brower & Egan,
1997; Brower & DeSalle, 1998).
The Extraction/PCR/Sequencing work was done over
a period of several years, so the procedures/protocols
has varied somewhat over this period. The methods
described here are the ones used most recently.
PHYLOGENY OF NYMPHALINI BUTTERFLIES
DNA extraction
Total DNA was extracted from adults using the abdomen (and later a single leg) or parts of larvae/pupae
using Qiagen’s QIAamp tissue kit (Qiagen GmbH,
Hilden, Germany) with the standard insect protocol.
The extracted DNA was stored at −20°C. Vouchers
have been stored at the Department of Zoology, Stockholm University.
Extractions were further purified with Qiagen’s QIAquick Spin PCR Purification Kit (Qiagen GmbH,
Hilden, Germany) before amplification. This improved
PCR efficiency.
PCR
PCR was performed on a Perkin Elmer Gene Amp
2400 using Amersham Pharmacia Biotech’s (Uppsala,
Sweden) Ready·To·Go PCR Beads (with 1 l of each
primer (at 10 mol/l) together with 3 l template DNA
and 20 l sterile distilled water). A typical cycling
profile for both nd1 and wingless was 95°C for 5 min
then 30×(95° for 30 s, 52° for 30 s, 72° for 1 min)
– hold at 4°. Some taxa needed variations to these
conditions. The amplified products were purified with
the QIAquick Spin PCR Purification kit (Qiagen
GmbH, 1993) and stored at −20°C until sequencing.
ND1
The mitochondrial nd1 gene codes for an NADH
subunit and is located between the 16S rRNA and
cytochrome b genes in the insect mitochondrion
(Clary & Wolstenholme, 1985). The 16S primer (5′TTCAAACCGGTGTAAGCCAGG-3′) of Weller et al.
(1994) was used in conjunction with a primer located
in the tRNA for Serine (5′-AAGCATTTGTTTTGAAAACTTAAG-3′) downstream from nd1. These two
primers amplify an 1155 bp section of mtDNA that
contains all of nd1.
Wingless
The wingless (wg) protein is a member of the wnt gene
family and is expressed at the wing margin in imaginal
discs in Drosophila playing a role in adult wing pattern
formation (Sidow, 1992; Carroll et al., 1994; Neumann
& Cohen, 1996).
PCR was performed using the primers of Brower
& DeSalle (1998): (LepWG1 (5′-GARTGYAARTGYCAYGGYATGTCTGG-3′), LepWG2 (5′-ACTICGCARCACCARTGGAATGTRCA-3′). These primers amplify
a 400 bp stretch of nuclear DNA in lepidopterans.
Sequencing
The double-stranded PCR-products were sequenced
using (Cy-5) labelled PCR-primers as well as internal
445
sequencing primers (Table 2) via the cyclic dideoxy
chain termination method using the Thermo Sequenase Fluorescent Sequencing kit from Amersham
Pharmacia Biotech AB, Uppsala, Sweden. A Corbett
thermal cycler was used with the following cycling
profile for nd1: 95° for 2 min – (95° for 30 s, 42° for
30 s, 70° for 1 min) × 30. And for wingless: 95° for
2 min – (95° for 30 s, 55° for 30 s, 70° for 1 min) × 30.
The reactions were electrophoresed on 6% Long Ranger
gels on the Pharmacia ALF-Express automated
sequencer. Both strands of the two genes were
sequenced (except for parts of nd1, due to sequencing
difficulties). The sequences were aligned using the
MacVector/AssemblyLIGN software/hardware package (International Biotechnologies, 1989). Almost no
indels and low base-pair divergence made alignment
uncomplicated.
PHYLOGENETIC ANALYSES
Monophyly of Nymphalini
In order to study the monophyly of Nymphalini with
respect to the close outgroup Kallimini a more distant
outgroup, Argynniti, was included in the analysis.
Subsequently we tested the effect of removing
Argynniti, to control for effects of a too distant
outgroup.
Representatives of the Melitaeini were not successfully sequenced, and Colobura dirce only for the
‘wingless’ gene. We performed an additional analysis
of ‘wingless’ which included also this species, using
only Argynnis paphia as outgroup.
The data sets
We used three main data sets, which were analysed
separately and together. The first data set, henceforth
referred to as the MEB data set (Appendices 1, 2),
contains 97 characters consisting of traits of morphology (including wing patterns), ecology and behaviour (including host plants). Ninety-six of these
traits are phylogenetically informative with respect to
the included taxa. The second data set consists of
695 mitochondrial nd1 characters (157 of which are
phylogenetically informative) and the third data set,
the nuclear wingless matrix, consists of 379 characters
(100 of which are phylogenetically informative). The
complete total evidence matrix consists of 31 taxa and
1171 characters, 353 of which are phylogenetically
informative. All characters are unweighted and unordered except for some characters from the MEB data
set which are ordered (see list of characters in Appendix
1). Gaps in molecular data were treated as a ‘fifth
base’, because indels were very rare and we wanted to
446
S. NYLIN ET AL.
Table 2. Primers employed
Gene
Primer name
Primer sequence (5′-3′)
PCR-primer (p)/internal
sequencing primer (i)
nd1
nd1
nd1
nd1
nd1
wingless
wingless
wingless
wingless
16s
nd1c
TrsLep2
nd440
ndr640
LepWG1
LepWG2
LepWG3
LepWG5
TTCAAACCGGTGTAAGCCAGG
TAGAATTAGAAGATCAACCAGC
AAGCATTTGTTTTGAAAACTTAAG
CAAACTATTTCTTATGAAGT
TCAGCAAAATCATAAGGAGT
GARTGYAARTGYCAYGGYATGTCTGG
ACTICGCARCACCARTGGAATGTRCA
ACIGIIAARACYTGYTGGATGAG
CGCARCACATRAGRTCGCARCCGTC
p
i
p
i
i
p
p
i
i
distinguish them from uncertain identification of
bases, coded as missing data.
Search options
The different data sets were analysed using PAUP∗
v 4b2 (Swofford, 1998). Most parsimonious trees were
constructed by heuristic searches using the TBR
branch swapping option and 10 000 random addition
sequences for the separate data sets and 100 000
for the total evidence data set. All analyses were
made with parsimony-uninformative characters excluded.
Support
Support for branches was investigated using decay
and bootstrap analysis. Decay indices were computed
with Autodecay 3.0.3 (Eriksson & Wikström, 1996)
using the PAUP search parameters: addseq=random,
nreps=1000, rseed=1.
Bootstraps were performed on Paup4b3a set to
5000 full heuristic (TBR) replicates with 10 random
addition sequences per bootstrap replicate. A
rearrangement limit of 10 000 000 rearrangements
per addition sequence was used for the wingless
data set as some bootstrap samples took too long to
compute otherwise. The rearrangement limit was not
hit, however, for all addition sequences in any single
bootstrap replicate.
Wing pattern and host plant data
For reasons explained in the Introduction, we performed a separate analysis of the data on adult wing
pattern (Wp 1-38 in Appendices 1, 2).
We were also interested in studying the degree of
dependence of the results on host plant data. For this
reason we performed an additional analysis of the
‘total evidence’ data set, with host plant data removed.
RESULTS
GENERAL
Decay indices and bootstrap values are given in Figures
1–4 to show support for specific branching patterns.
Parsimony analysis of the data set based on morphology, ecology and behaviour (including host plants;
Table 2; the MEB data set) resulted in eight most
parsimonious trees. A strict consensus tree is shown
in Figure 1.
Parsimony analysis of the mitochondrial data set
(nd1) resulted in one most parsimonious tree, shown
in Figure 2. Note the weak phylogenetic signal in this
data set, as indicated by low decay values and bootstrap
support, little resolution and low consistency index
(0.365).
Parsimony analysis of the ‘wingless’ data set resulted
in 72 most parsimonious trees, a strict consensus tree
is shown in Figure 3. The consistency index (but not
retention index) of this data set (0.530) is higher than
for either MEB (0.431) or nd1 (0.365).
Figure 4 shows the strict consensus tree for the two
most parsimonious trees found by analysis of total
evidence. These trees have a consistency index of
0.402.
MONOPHYLY OF NYMPHALINI
Nymphalini did not appear as a monophyletic group
in the analysis based on the MEB data set (Fig. 1). In
this analysis Mynes and Symbrenthia ended up with
members of the Kallimini in the outgroup. However,
nd1 data (Fig. 2) as well as ‘wingless’ data (Fig. 3)
support a monophyletic Nymphalini, and so does total
evidence (Fig. 4).
Removing the distantly related Argynniti from the
outgroup did not change this result, and had very
little effect on branching patterns (Antanartia moves
from its position with the Vanessa-type genera (Fig.
4) to being the sister taxon to the focal group,
PHYLOGENY OF NYMPHALINI BUTTERFLIES
447
Argynniti
P. coenia
1
27
H. bolina
1
30
7
95
M. geoffroyi
Symbrenthia sp.
A. levana
Hypanartia sp.
2
39
A. schaeneia
1
1
37
59
2
V. atalanta
53
B. gonerilla
1
52
C. cardui
C. virginiensis
I. io
2
48
1
37
2
38
8
98
2
40
2
56
5
90
A. urticae
5
86
N. antiopa
A. milberti
N. cyanomelas
N. xanthomelas
N. polychloros
N.californica
R. l-album
K. canace
6
91
3
91
2
84
1
80
1
66
P. interrogationis
P. comma
P. c-aureum
P. satyrus
P. progne
6
97
P. gracilis
1
73
P. faunus
P. c-album
P. egea
Figure 1. Results from cladistic analysis of data on morphology, ecology and behaviour (the MEB database). Shown
is a strict consensus tree of eight most parsimonious trees (311 steps) with consistency index (CI) 0.431 and retention
index (RI) 0.747. Decay indices (top) and boot strap values (bottom) on branches show level of support.
Polygonia, Kaniska, Nymphalis, Roddia, Aglais and
Inachis).
In the separate analysis of ‘wingless’ data, which also
included C. dirce, this representative of the Coloburini
ended up together with the outgroup, A. paphia.
Nymphalini was again monophyletic, with the
Kallimini as sister group. Note, however, that the tribe
Melitaeini was not represented.
MAJOR CLADES IN NYMPHALINI
It should be noted that there is some agreement between phylogenies resulting from the three data sets
regarding several aspects of the basic topology (Figs
1–3). Hypanartia is given a basal position by both nd1
and ‘wingless’ data (Figs 2, 3), and this is the position
favoured by total evidence (Fig. 4), although MEB data
448
S. NYLIN ET AL.
H. bolina
3
P. coenia
26
Argynniti
Hypanartia sp.
M. geoffroyi
1
17
4
30
3
13
1
8
2
26
A. levana
2
22
R. l-album
6
84
A. milberti
1
<5
Symbrenthia sp.
K. canace
A. urticae
A. schaeneia
I. io
1
<5
V. atalanta
3
24
2
<5
1
21
2
<5
4
48
3
17
3
61
4
60
1
26
3
4
89
40
1
11
C. cardui
C. virginiensis
N.californica
N. xanthomelas
N. polychloros
N. antiopa
P. egea
P. c-aureum
P. faunus
P. c-album
P. interrogationis
3
53
P. comma
6
90
P. progne
2
72
6
99
P. satyrus
P. gracilis
Figure 2. Results from cladistic analysis of data from the nd1 gene. Shown is a single most parsimonious tree (631
steps) with CI 0.365 and RI 0.417. Decay indices (top) and boot strap values (bottom) on branches show level of support.
suggest a position with the Vanessa-type genera and
Antanartia (Fig. 1).
Another basal clade supported by total evidence
consists of Mynes, Symbrenthia and Araschnia, the
first two being sister genera (Fig. 4). This arrangement
is also seen after analysis of the ‘wingless’ data alone
(Fig. 3), whereas MEB data joins Mynes and Symbrenthia but not Araschnia (Fig. 1) and nd1 data joins
the three genera but in a different arrangement (Fig.
2). The support for placing these three genera outside
of the Vanessa-type genera and the focal group is rather
weak (Figs 1–4), but the consistency between the types
of data speak in favour of it.
Approaching the focal group, there is moderate support, in the analysis of total evidence, for a clade
corresponding to Vanessa in the wider sense (Vanessa
+ Cynthia + Bassaris) and weak support for also
placing Antanartia on this branch (Fig. 4). MEB data
PHYLOGENY OF NYMPHALINI BUTTERFLIES
449
Argynniti
2
P. coenia
73
H. bolina
Hypanartia sp.
V. atalanta
2
3
53
62
C. cardui
C. virginiensis
A. levana
2
47
1
30
2
65
M. geoffroyi
Symbrenthia sp.
B. gonerilla
A. schaeneia
1
38
1
96
29
I. io
5
4
100
A. milberti
A. urticae
K. canace
2
50
R. l-album
N. polychloros
1
73
N.californica
1
4
76
97
N. antiopa
N. xanthomelas
P. c-aureum
P. egea
1
64
1
60
1
83
P. faunus
3
94
P. gracilis
P. c-album
P. progne
P. comma
1
78
P. interrogationis
P. satyrus
Figure 3. Results from cladistic analysis of data from the ‘wingless’ gene. Shown is a strict consensus tree of 72 most
parsimonious trees (270 steps) with CI 0.530 and RI 0.601. Decay indices (top) and boot strap values (bottom) on
branches show level of support.
alone suggest a larger clade including also Hypanartia
and Araschnia (Fig. 1). The phylogeny suggested by
nd1 data joins Vanessa with Cynthia but is otherwise
unresolved (Fig. 2; Bassaris was not successfully
sequenced) and ‘wingless’ gives even less information
(Fig. 3).
There is strong support by total evidence for a predominantly holarctic major clade corresponding to the
focal group, consisting of Polygonia, Kaniska, Nymphalis, Roddia, Aglais and Inachis (Fig. 4). This clade
has strong support in the MEB data analysis (Fig. 1),
and is suggested also by ‘wingless’ data (although
450
S. NYLIN ET AL.
Argynniti
2
60
P. coenia
H. bolina
Hypanartia sp.
A. levana
5
64
9
90
10
94
M. geoffroyi
Symbrenthia sp.
A. schaeneia
1
20
2
36
B. gonerilla
3
67
2
31
1
14
V. atalanta
4
68
C. cardui
C. virginiensis
I. io
8
98
13
A. urticae
100
A. milberti
N. polychloros
9
3
78
99
N.californica
2
73
4
84
1
54
N. xanthomelas
5
94
N. antiopa
N. cyanomelas
R. l-album
K. canace
2
62
P. egea
1
5
92
40
15
100
11
99
P. faunus
P. c-album
P. interrogationis
P. c-aureum
1
31
P. comma
3
56
P. satyrus
3
63
2
66
P. progne
P. gracilis
Figure 4. Results from cladistic analysis of total evidence. Shown is a strict consensus tree of two most parsimonious
trees (1272 steps) with CI 0.402 and RI 0.553. Decay indices (top) and boot strap values (bottom) on branches show
level of support.
invaded by Antanartia; Fig. 3). Nd1 data give little
evidence for or against this clade (Fig. 2).
GENERIC RELATIONSHIPS WITHIN THE FOCAL GROUP
Apparently there are two major clades in the focal
group, one consisting of Aglais + Inachis, the other
of the remaining genera. A close relationship between
Aglais and Inachis is supported by ‘wingless’ data (Fig.
3) and by analysis of total evidence (Fig. 4). Once
again, nd1 contributes little information (Fig. 2) but
it should be noted that this is another data set that
gives no support to the common practice of including
Aglais in Nymphalis. MEB data weakly joins Aglais
PHYLOGENY OF NYMPHALINI BUTTERFLIES
and Inachis with Nymphalis (Fig. 1) but a position
outside of the other genera in the focal group is suggested by ‘wingless’ data (Fig. 3) and partly by nd1
data (Fig. 2). This position is favoured with moderate
support by total evidence (Fig. 4).
The bulk of the species in the remaining genera
belongs to either Nymphalis or Polygonia in the strictest sense. Kaniska, the only species of which (canace)
is often placed in Polygonia, is most probably the sister
taxon to this genus. This is suggested by total evidence
(Fig. 4), due mostly to MEB data (Fig. 1). Molecular
data do not resolve relationships among genera in this
group (Figs 2, 3).
Regarding relationships between genera, the most
problematic is the position of the monotypic Roddia
(a.k.a. ‘Nymphalis vaualbum’). The only species in
this genus has been placed in either Nymphalis or
Polygonia by various authors. MEB data rather
strongly suggest a position with Kaniska and Polygonia
(Fig. 1), and this position is favoured by total evidence
as well (Fig. 4). Nd1 data weakly join Roddia with
Kaniska, but outside of both Nymphalis and Polygonia
(Fig. 2). ‘Wingless’ data weakly put Roddia with
Nymphalis (Fig. 3).
451
closest associate (Fig. 2). In contrast, ‘wingless’ data
join P. gracilis with P. progne, and as this arrangement
is also supported by some of the individual trees behind
the consensus tree in Figure 1, it is the one that occurs
after analysis of total evidence (Fig. 4). The two most
parsimonious trees obtained in the analysis of total
evidence conflict only regarding the position of P. caureum, inside or outside of P. interrogationis. The
outside position is the one favoured after successive
weighting (Fig. 4).
WING PATTERN DATA
We performed a separate analysis with only the wing
pattern data, which resulted in 180 most parsimonious
trees, the strict consensus of which is seen in Figure
5. The consistency index was as high as 0.515. Note
that the topology found is broadly similar to that
supported by molecular data (Figs 2, 3) and total
evidence (Fig. 4). Mynes and Symbrentia are joined in
a basal position, Hypanartia is found outside of the
other Vanessa-like genera (which are joined), the focal
group and the genus Polygonia in the narrow sense
are monophyletic, and the relationships within Polygonia are relatively similar.
SPECIES RELATIONSHIPS IN THE FOCAL GROUP
EFFECTS OF HOST PLANT DATA
Species relationships within Nymphalis and Polygonia
were not resolved in this study. Within Nymphalis,
antiopa and cyanomelas are evidently sister taxa, but
this is based only on a subset of the MEB data matrix
(Fig. 1) and no molecular data are present for N.
cyanomelas. The three data sets conflict regarding
the relative position of the remaining species. Total
evidence favours the ladder arrangement seen in
Figure 4, with N. polychloros as the basal species, but
only with moderate support. The basal position for N.
polychloros is the result of ‘wingless’ data (Fig. 3),
conflicting arrangements have weaker support from
other kinds of data (Figs 1, 2).
Within Polygonia the only strongly supported relationship is that P. c-album and P. faunus are sister
taxa (Fig. 4). This clade appears in the separate
analyses of all three data sets (Figs 1–3). Total
evidence weakly supports association of the western
palearctic P. egea with this clade, and a major clade
consisting of the remaining nearctic species and the
eastern palearctic P. c-aureum (Fig. 4). Within the
latter group there is moderate support for a clade
consisting of P. comma, P. satyrus, P. gracilis and P.
progne (Fig. 4).
The phylogeny based on MEB data is unresolved
concerning these relationships (Fig. 1). Nd1 data show
weak support for the last-mentioned clade of four
species and strongly support a close relationship between P. satyrus and P. gracilis, with P. progne as the
The analysis of total evidence without the host plant
data character (ec2) resulted in two most parsimonious
trees with a consistency index of 0.398. The consensus
tree was identical to the one obtained previously from
total evidence (Fig. 4).
DISCUSSION
MONOPHYLY AND POSITION OF NYMPHALINI
Harvey (1991) recognizes 13 subfamilies of the Nymphalidae, among them the Nymphalinae. This subfamily
is further divided into the three tribes Nymphalini,
Melitaeini and Kallimini. In contrast, Ackery (1988)
recognized a larger Nymphalini, corresponding to Harvey’s Nymphalini + Kallimini, and also the tribe Coloburini as members of the Nymphalinae, but not the
Melitaeini. Harvey notes that Kallimini and Melitaeni
are united by one larval trait, the presence of filiform
seta on the sclerotized base of the scolus on A9. On
the other hand, Nymphalini and Melitaeini are united
by the presence of filiform setae on A1,2. Coloburini
lack both of these traits, according to Harvey (1991).
These traits were not studied by us. Overall, taking
also host plants into account, Harvey favoured the
hypothesis of a sister-group relationship between
Kallimini and Melitaeini, Nymphalini being the sistergroup to this clade.
The synapomorphies used by Harvey (1991) to define
452
S. NYLIN ET AL.
Argynniti
P. coenia
H. bolina
A. levana
A. schaeneia
1
C. cardui
C. virginiensis
V. atalanta
B. gonerilla
Hypanartia sp.
1
1
1
I. io
N. antiopa
N. cyanomelas
K. canace
A. urticae
A. milberti
N. polychloros
N.californica
N. xanthomelas
1
R. l-album
1
P. interrogationis
P. comma
1
P. satyrus
1
1
P. c-aureum
1
P. faunus
P. c-album
4
P. egea
P. progne
P. gracilis
2
M. geoffroyi
Symbrenthia sp.
Figure 5. Results from cladistic analysis of only wing pattern traits (38 characters). Shown is a strict consensus tree
of 180 most parsimonious trees (103 steps) with CI 0.515 and RI 0.811.
the Nymphalini as distinct from the Kallimini and
Melitaeini are the hairy eyes (our character ey1) and
the stiff, projecting bristlelike scales on the palpi. He
noted that the latter trait is absent in Mynes and that
projecting, flat scales are present in some Kallimini.
There was no further description of this trait, and we
failed to see a clear difference between members of
Nymphalini and Kallimini. Our traits lp1-4 deal with
the morphology of palpi. We found eyes to be hairy in
all studied species of the ingroup, and not in any of
the outgroup species.
Our analysis was not specifically designed to answer
most of the questions regarding relationships among
tribes (this will be done in a companion study), but
our results suggest monophyly of Nymphalini sensu
(Harvey, 1991), relative to the Kallimini and Coloburini
PHYLOGENY OF NYMPHALINI BUTTERFLIES
(the latter conclusion based on the analysis of ‘wingless’
data). Melitaeini was not included in the analysis, but
this is a uniform tribe very likely to be monophyletic
and hence it is unlikely that members of this tribe
would intrude in the Nymphalini.
PHYLOGENY OF NYMPHALINI AT THE GENUS LEVEL
We found some support for a basal position of the
predominantly neotropical genus Hypanartia. This
conflicts with the present view of a close relationship
between this genus and the African Antanartia (DeVries, 1987). We are not aware of any proposed synapomorphies joining these two genera, and our results
suggest that the similarities between them may be
plesiomorphic. A position for Hypanartia together with
Antanartia results in a tree which is four steps longer,
a position inside of the Mynes + Symbrenthia +
Araschnia clade three steps longer, and all other positions (including instead moving Antanartia to the
Hypanartia branch) result in considerably longer trees.
The position of Hypanartia should be considered as
tentative, as it is not supported by any uniquely derived
traits in our matrix. Seven traits in the MEB data
change along the branch leading to the remaining
Nymphalini, but none of them unequivocally, and they
all change again higher in the tree. Seven molecular
traits change unequivocally along this branch, but all
of them change again. However, we did not have access
to the juvenile stages of Hypanartia. It should be noted
that Müller (1886) on the basis of juvenile morphology
divided his ‘Vanessinae’ into two groups; one group
consisting of only Hypanartia, the other of his ‘Pyrameis’, ‘Vanessa’ and ‘Grapta’. Scrutiny of the studied
species reveals that he considered members of modern
Cynthia, Vanessa, Araschnia, Aglais, Inachis, Nymphalis and Polygonia to all belong to the second group!
He states that it is especially the smooth, compressed
pupa that sets Hypanartia apart from the other genera.
Also unexpectedly, we found consistent support for
another basal clade consisting of Mynes, Symbrenthia
and Araschnia, with strong support for a sister group
relationship between the two former genera. A basal
position for Araschnia relative to the Vanessa- and
Nymphalis-type species has been suggested by Niculescu (1965, 1985) and Teshirogi (1990). These
authors studied only palearctic genera. One
synapomorphy for this clade of three genera is apparently the absence of sclerotized spines on larval
abdominal segment 9 (the narrow penultimate segment), where two subdorsal spines occur in last-instar
larvae of the remaining genera studied. Teshirogi
(1990) noted the absence of spines on this segment
in Araschnia levana, A. burejana and Symbrenthia
hippoclus, and we found them to be absent in Mynes
geoffroyi as well. Müller (1886) and Niculescu (1965)
453
note the presence of spines on all abdominal segments
in A. levana, apparently mistakenly (like Teshirogi,
(1990), we found them to be absent on this segment,
also in a European stock). A remaining uncertainty
concerns the state of this trait in Hypanartia, as we
could not study larval material of this species ourselves.
Another potential synapomorphy for the clade is
the presence, in first-instar larvae, of several short
secondary setae on the small pinaculae on which P1
and Sp1 arise on abdominal segment 10. These were
noted by Nakanishi (1988) to occur in Symbrenthia
and Araschnia and to be absent in all other genera in
Nymphalini and Kallimini studied by him (which led
him to suggest a sister-group relationship). However,
first instar larvae of Mynes do not seem to have been
studied, and we did not have access to early-instar
larval material. In addition, in the molecular data
there are three synapomorphies, i.e., traits changing
uniquely and unequivocally along the branch leading
to these three genera. Moving Araschnia to the clade
of Vanessa-type genera, as might be suggested by the
MEB data (wing pattern similarity), results in a tree
that is seven steps longer.
The position for this clade of three genera outside
of both the Vanessa-type genera and the focal group is
only weakly supported (Fig. 4). One character in support of this arrangement is the hibernating stage (ec1),
which is the adult stage in both the Vanessa-group
and in the focal group (some taxa lack hibernation
diapause), whereas a juvenile stage (egg, larva or pupa)
is used in those temperate genera placed outside by
total evidence. Adult hibernation is rare in butterflies,
and this potential synapomorphy should be given some
consideration.
A close relationship between Mynes and Symbrenthia has to our knowledge not been suggested
previously. Support for this clade was high from the
MEB data (Fig. 1) but there was no synapomorphy
without homoplasy in this data set. Two synapomorphies were suggested by molecular data, one
of them an unequivocal change along this branch. The
alternative arrangement with Mynes outside of the
other two genera, suggested by nd1 data (Fig. 2) results
in a tree which is no less than 11 steps longer.
As expected, we found some support for a close
relationship between the genera Vanessa, Cynthia and
Bassaris, sensu Field (1971). Many subsequent authors
have not followed Field’s division into separate genera.
With the limited number of species included in this
analysis our results have no bearing on the question of
whether the three genera can be defended on cladistic
principles, i.e. whether they are all monophyletic. Interestingly, the association between them is not so
strongly supported by the analysis as could have been
454
S. NYLIN ET AL.
expected. We found no good candidates for synapomorphies in the molecular data or in the MEB
data, despite the obvious similarities. The apparently
basal position of Hypanartia, often considered to be
another Vanessa-type genus, suggests the possibility
that most of the similarities are in fact plesiomorphic.
As Field (1971) did find diagnostic differences between
the genera, it may be prudent to recognize them as
separate, at least until there is better evidence that
they really are a closely related monophyletic group.
We believe that we can settle with some confidence
a couple of long-standing controversies regarding the
taxonomy of butterflies in the Nymphalini. The first
regards the position of the palearctic A. urticae and
the nearctic A. milberti. The former has long been
placed in the genus Aglais by Eurasian authors, but
the latter is often placed in Nymphalis by American
and Canadian authors (e.g., Scott, 1986; Layberry et
al., 1998). However, it has long been recognized that
the two species are closely related (Scudder, 1889;
Seitz, 1914; Miller & Miller, 1990), as confirmed by
the present study. We show here that the practice of
placing these species in Nymphalis cannot be defended
from cladistic principles, as Aglais is in fact most
closely related to Inachis, and Nymphalis is more
closely related to Polygonia and relatives. The same
conclusion was reached by Niculescu (1965) and
Teshirogi (1990). The similarity between Aglais and
Nymphalis is plesiomorphic and superficial, as demonstrated also by the fact that the male genitalia are
very different (Niculescu, 1965).
Aglais and Inachis are united by their shared habit
of laying very large batches of eggs, in many layers,
on the underside of nettle (Urtica) leaves. In the MEB
data set this complex character was split into several
traits (batch size, site and shape, host plant) in order
to maximize the number of potentially informative
characters, and none of them change uniquely and
unequivocally along the branch leading to the two
genera. However, no less than five molecular traits do
so. Three of them consists of adjacent bases in the
‘wingless’ data which are found to be missing in both
(and only) I. io, A. urticae and A. milberti when aligning
the sequences. This might be considered a single trait,
but as gaps are very infrequent in the studied sequences we chose to give this extraordinary evolutionary event added weight by treating it as three
separate traits in the analyses.
The position of Aglais + Inachis outside of the
remaining genera in the focal group has moderately
strong support (Fig. 4). There are no traits in the MEB
database uniquely uniting the other genera, but one
change in the molecular data does so. The alternative
position with best support is as the sister clade to
Nymphalis; however, this tree is four steps longer.
The second controversy has regarded the position of
R. l-album and K. canace, which are shown here to most
probably belong in the Polygonia clade. The proposed
position of R. l-album with Polygonia, suggested earlier
by e.g. Niculescu (1965, 1985), is supported by total
evidence (Fig. 4), albeit not very strongly. Niculescu
relied partly on the presence of larval head horns in
R. l-album and Polygonia (absent in Nymphalis) but
such horns are absent in K. canace as well (our trait
L2). There are four other potential synapomorphies
in the MEB database joining Roddia, Kaniska and
Polygonia. These are: lp1, the palpi have a distinctly
set off apical segment; ws2, the posterior part of the
fore wing is already more incised in R. l-album than
in Nymphalis, approaching the state in canace and
Polygonia; ws5, the outer margin of the fore wing is
deeply incised; and wp3, the presence of an angled
white spot on the underside of the hind wings which
has given the ‘commas’ (Polygonia) their common
name, and (ironically) the epithet ‘The false comma’ to
R. l-album. The alternative position, with Nymphalis,
also has supporting traits, e.g. the slender shape of
the antennal club (a1) and the yellowish colour of the
eggs (e5); however, this tree is three steps longer.
Teshirogi (1990) arrived at a similar conclusion, favouring a closer relationship with Polygonia but noting
the remaining uncertainty. He also did not resolve
the relationship between R. l-album, K. canace and
Polygonia in the narrow sense, whereas we see the
closer relationship between the latter two taxa as
relatively certain. There is relatively strong support
for this clade (Fig. 4) and there are two synapomorphies
in the MEB database (ws2, the deeply incised posterior
part of the fore wing; and L7, the particular coloration
seen in larva) and one in the molecular data.
Polygonia in the narrow sense forms a very welldefined clade with high support (Fig. 4). This is probably one reason for the reluctance by many authors to
include the two controversial taxa in Polygonia. This
is particularly true in the case of R. l-album, which in
many respects resembles the species in Nymphalis
more than it resembles the species in Polygonia s.s..
This similarity is evidently plesiomorphic and if so
cannot form the basis for taxonomic groupings. However, a taxonomy could be adopted which recognizes
the many traits which distinguish Polygonia s.s., as
well as the conflicting evidence regarding the position
of R. l-album, which leaves this species as something
of an ‘evolutionary link’ between Nymphalis and Polygonia. The differences between the narrow sense Polygonia and K. canace in e.g., adult coloration and larval
host plants are certainly impressive enough to warrant
use of the genus Kaniska for the latter. If so, R. lalbum (which is even more distantly related to the
narrow sense Polygonia) must also be placed in a genus
of its own. The appropriate name for this genus would
PHYLOGENY OF NYMPHALINI BUTTERFLIES
seem to be Roddia, recently suggested by Korshunov
(Korshunov & Gorbunov, 1995; Korshunov, 1996).
PHYLOGENY OF THE NYMPHALIS GROUP AT THE
SPECIES LEVEL
Relationships within Nymphalis remain uncertain, except for the well-supported clade N. antiopa + N.
cyanomelas, suggested earlier by Scudder (1889) and
Miller & Miller (1990). The high support may partly
be the result of lack of conflicting evidence, due to
absence of molecular data and juvenile traits for N.
cyanomelas.
Miller & Miller (1990) suggested the clade N. californica + N. xanthomelas, based on similarity of male
genitalia. This is worth noting because the clade is
supported also by our nd1 data (Fig. 2). However,
we failed to see any unique similarities between the
genitalia of the two species from the illustrations in
Miller & Miller (1990), and there was no description
given of the alleged similarity. In the absence of synapomorphies, the ‘similarity’ (and possibly the nd1
results) is equally well explained by the ladder arrangement in Figure 4, and hence we provisionally
favour the resolution supported by total evidence.
The basal position for N. polychloros is relatively
well supported by ‘wingless’ data (Fig. 3), including
one synapomorphy for the remaining species (N. cyanomelas not sequenced).
PHYLOGENY OF POLYGONIA AT THE SPECIES LEVEL
As noted in the Results section, the resolution within
Polygonia at the species level supported by total evidence (Fig. 4) must be considered as tentative (except for
faunus + c-album), although it is the best hypothesis
presently available. The sister-group relationship between P. faunus and P. c-album is supported by e.g.
larval coloration (L6) and two synapomorphies in the
molecular data. Adults are also very similar, and larvae
have similar polyphagous habits. These traits were
not well captured by the particular coding that we
employed, but nevertheless, support for this clade was
one of the highest in the phylogeny.
The nearctic clade P. comma + P. progne + P. gracilis
+ P. satyrus is not supported by any synapomorphies.
One molecular synapomorphy supports monophyly of
the latter three genera. In our earlier analyses (with
less data) P. comma grouped instead with P. interrogationis, and more data are needed before we can
conclusively choose between these alternatives. The
same is true for the relationships between progne,
gracilis and satyrus, where nd1 shows very strong
similarity between the last two species, conflicting
with total evidence.
455
WING PATTERN DATA
Analysis of only wing pattern characters resulted in a
topology broadly similar to the one resulting from
analysis of the complete matrix or only molecular data
(Fig. 5, cf. Figs 1–4). This suggests that wing pattern
data can be very useful in analysing butterfly relationships, a fact that may be of importance considering that such characters can be collected easily
and non-destructively from museum material, and if
necessary even from good illustrations.
There are pitfalls, however, illustrated by the clearly
artificial grouping of Nymphalis antiopa (and cyanomelas) with I. io and K. canace. These species
evidently have lost much of their ancestral pattern
(the one seen in Aglais, Roddia, Polygonia and other
Nymphalis), and for this reason lack the specific synapomorphies joining other species. Consequently they
tend to be pulled together by superficial similarities.
Another example is the position of Araschnia with the
Vanessa-type genera (Fig. 5), which is only supported
by wing pattern similarities that may well be plesiomorphic for the entire tribe. Clearly, other kinds of
data will often be needed as well, in order to obtain a
reliable hypothesis of phylogeny.
PHYLOGENETIC ECOLOGY – INCLUDE THE STUDY
TRAITS OR NOT?
Phylogenies are used in many areas of evolutionary
biology, e.g. ecology and ethology. This is either because
a detailed historical reconstruction is necessary to test
or suggest theories about the evolutionary process, or
in order to control for the phylogenetic interdependence
of species characteristics in statistical tests (Wanntorp
et al., 1990; Brooks & McLennan, 1991; Harvey &
Pagel, 1991; Miller & Wenzel, 1995). When the use of
such methods first became popular in the early 1990s,
it was suggested that it was important that the traits
under study, e.g. the ‘ecological’ characters were not
used to construct the phylogeny in the first place
(Brooks & McLennan, 1991). This is because this procedure would induce some circularity; the number of
transformations in the studied traits would be underestimated because branches in the phylogeny with
the same state would more often be placed together in
the search for the most parsimonious arrangement of
the taxa.
This conclusion has, however, been debated (see
Miller & Wenzel, 1995 for additional references) and
recently Zrzavy (1997) argued strongly that all available characters should be included in the data matrix.
He reasoned that problems of circularity should in fact
be less if, for example, ecological traits are included
rather than excluded, because the null hypothesis
should be that the ecological traits are historically
contingent and so an adaptive explanation is not
456
S. NYLIN ET AL.
needed for each state in each taxon. The null hypothesis
is equally strong or stronger when ecological traits are
included, so this procedure is more conservative and
does not overestimate adaptive change.
We agree in principle with this reasoning, and so
have included host plant data in the data matrix,
although we intend to use the phylogeny to study
butterfly–plant relationships (Janz, Nyblom & Nylin,
2001). However, some of the uses that we intend for
the phylogeny have particular problems in this respect.
These problems arise when the question under study
is not the typical one in phylogenetic ecology (whether
some environmental variable has affected a biological
characteristic, or one biological trait has affected another during evolution) but instead the question concerns the transformation rates themselves. For
instance: is host plant use in butterflies phylogenetically constrained and, if so, in what ways (Janz
& Nylin, 1998)? Including host plants in an otherwise
small data matrix will tend to result in a reconstruction
showing a conservative host plant utilization, whether
this is the real situation or not.
In our opinion, whenever the transformation rates
themselves are of interest it is necessary to include
the traits under study in the data matrix, but also to
show that the resulting tree is stable enough to show
essentially the same topology if the study traits are
subsequently excluded. Phylogenetic ecologists will
typically not go to the trouble of testing whether the
phylogenetic hypothesis from the literature that they
are using would change if their study traits were
included in the data matrix, but chances are that it
would, judging from our experiences in the present
study. We found that our results were very dependent
on inclusion of host plant data, even with a large
morphological and ecological data matrix, as well as
when we added data from the nd1 gene. Not until we
added also data from the ‘wingless’ gene did we have
a large enough data matrix to achieve a tree stable
against inclusion/exclusion of host plant data. Moreover, the chances of finding the tree conforming to the
‘true’ phylogeny will be better with more data. Using
a published tree based on, for example, a single gene,
without combining these data with the researcher’s
database on ecological traits will often amount to using
a phylogeny which is not the best hypothesis available.
ACKNOWLEDGEMENTS
We are very grateful to all those who in various ways
helped us obtaining the specimens and/or with important information: Phil Ackery, Yuri. Iv. Berezhnoi,
Michael Braby, Andrew Brower, Steve Collins, Carlos
Cordero, Jürg DeMarmels, Konrad Fiedler, Mecky
Furr, Enrique Garcia-Barros, Dianne Gleeson, Karl
Gotthard, Cris Guppy, Bert Gustafsson, Henry Hensel,
Darrell Kemp, Norbert Kondla, James Kruse, Jaakko
Kullberg, Dave Lohman, Pat Lorch, Armando Luis,
Christopher Majka, Don Miller, Adolfo Navarro, Guy
van de Poel, David Pollock, James Scott, Kojiro Shiraiwa, Masao Taguchi, Toomas Tammaru, John & Jill
Thompson, Aki Tsuneda, Niklas Wahlberg, Mamuro
Watanabe, Wayne Wheeling, Per-Olof Wickman, Christer Wiklund, Myron Zalucki, Cor Zonneveld, and many
others. This research was supported by grants from
the Swedish Natural Science Research Council to S.N.
REFERENCES
Ackery PR. 1988. Hostplants and classification: a review of
nymphalid butterflies. Biological Journal of the Linnean
Society 33: 95–203.
Aubert J, Barascud B, Descimon H, Michel F. 1996.
Systématique moléculaire des Argynnes (Lepidoptera:
Nymphalidae). Compte Rendue de l’Academie des Sciences,
Paris, Life Sciences 319: 647–651.
Brooks DR, McLennan DH. 1991. Phylogeny, ecology, and
behavior. A research program in comparative biology. Chicago: University of Chicago Press.
Brower AVZ, DeSalle R. 1998. Patterns of mitochondrial
versus nuclear DNA sequence divergence among nymphalid butterflies: the utility of wingless as a source of characters for phylogenetic inference. Insect Molecular Biology
7: 73–82.
Brower AVZ, Egan MG. 1997. Cladistic analysis of Heliconius butterflies and relatives (Nymphalidae: Heliconiiti):
a revised phylogenetic position for Eueides based on sequences from mtDNA and a nuclear gene. Proceedings of
the Royal Society of London B 264: 969–977.
Carroll SB, Gates J, Keys DN, Paddock SW, Panganiban
GEF, Selegue JE, Williams JA. 1994. Pattern formation
and eyespot determination in butterfly wings. Science 265:
109–114.
Clary DO, Wolstenholme DR. 1985. The mitichondrial
DNA molecule of Drosophila yakuba: nucleotide sequence,
gene organization and genetic code. Journal of Molecular
Evolution 22: 252–271.
Common IFB, Waterhouse DF. 1972. Butterflies of Australia. Sydney: Angus and Robertson.
De la Maza EJ, White Lopez A. 1986. Redescubrimento
de Nymphalis cyanomelas (Dbld. & Hew.) en Mexico.
(Nymphalidae: Nymphalinae). Revista de la Sociedad Mexicana de Lepidopterologica X: 35–39.
DeVries PJ. 1987. The Butterflies of Costa Rica and their
Natural History: Papilionidae, Pieridae, Nymphalidae.
Princeton: Princeton University Press.
Eriksson T, Wikström N. 1996. Auto Decay. Version 3.0.3
ed. Stockholm, Sweden: Department of Botany, Stockholm
University (distributed by the authors).
Esper EJC. 1780. Die Schmetterlinge in Abbildungen nach
der Natur mit Beschreibungen von Eugenius Johann
Cristoph Esper. Erste Theil. Europäische Gattungen. Leipzig: L.D. Weigel.
Field WD. 1971. Butterflies of the genus Vanessa and of
PHYLOGENY OF NYMPHALINI BUTTERFLIES
the resurrected genera Bassaris and Cynthia (Lepidoptera:
Nymphalidae). Washington: Smithsonian Institution
Press.
Harvey DJ. 1991. Higher classification of the Nymphalidae.
In: Nijhout HF, ed. The development and evolution of butterfly wing patterns. Washington: Smithsonian University
Press, 255–276.
Harvey PH, Pagel MD. 1991. The comparative method in
evolutionary biology. Oxford: Oxford University Press.
Hawkswood TJ. 1990. Observations on the egg of Mynes
geoffroyi guerini Wallace (Lepidoptera: Nymphalidae). Victorian Entomologist 20: 148–150.
Higgins LG. 1975. The classification of European butterflies.
London: Collins.
Janz N, Nyblom K, Nylin. 2001. Evolutionary dynamics of
host-plant specialization: a case study of the tribe Nymphalini. Evolution 58: 783–796.
Janz N, Nylin S. 1998. Butterflies and plants: a phylogenetic
study. Evolution 52: 486–502.
Johnston G, Johnston B. 1980. This is Hong Kong: Butterflies. Hong Kong: Hong Kong Government Publications.
Kocak AÖ. 1981. Critical check-list of European Papilionoidea. Priamus 1: 46–90.
Korshunov Y. 1996. Dopolneniya i ispravleniya k knigen
“Dnevnye babochki aziatskoi chasti Rossii”. [Additions and
corrections to the book “Butterflies of the Asian part of
Russia”]. Novosibirsk: ETA Group.
Korshunov Y, Gorbunov P. 1995. Dnevnye babochki aziatskoi chasti Rossii. Spravochnik. [Butterflies of the Asian
part of Russia. A handbook]. Ekaterinburg: Ural University
Press.
Larsen TB. 1991. The butterflies of Kenya and their natural
history. Oxford: Oxford University Press.
Layberry RA, Hall PW, Lafontaine JD. 1998. The butterflies of Canada. Toronto: University of Toronto Press.
Miller JS, Wenzel JW. 1995. Ecological characters and
phylogeny. Annual Review of Entomology 40: 389–415.
Miller LD, Miller JY. 1990. Nearctic Aglais and Nymphalis
(Lepidoptera, Nymphalidae): Laurasia revisited? The Entomologist 109: 106–115.
Müller W. 1886. Südamerikanische Nymphalidenraupen:
Versuch eines natürlichen Systems der Nymphaliden. Zoologische Jahrbücher, Abteilung Systematik 1: 417–678.
Nakanishi A. 1988. Study on the first instar larvae of
the subfamily Nymphalinae (Lepidoptera, Nymphalidae).
Special Bulletin of Lepidopterological Society of Japan 6:
83–99.
457
Neumann CJ, Cohen SM. 1996. Distinct mitigenic and cell
fate specification function of wingless in different regions
of the wing. Development 122: 1781–1789.
Niculescu EV. 1965. Familia Nymphalidae. Fauna Republicii Populare Romane 11. Bucharest: Editura Academiei Republicii Populare Romane.
Niculescu EV. 1985. Problèmes de systématique dans la
famille des Nymphalidae. Deutsche Entomologische Zeitschrift, NF 32: 335–347.
Nijhout HF. 1991. The development and evolution of butterfly
wing patterns. Washington: Smithsonian University Press.
Scott JA. 1986. The butterflies of North America. Stanford,
CA: Stanford University Press.
Scudder SH. 1889. The butterflies of the Eastern United
States and Canada with special reference to New England.
Cambridge: Published by the author.
Seitz A. 1914. Genus Vanessa. In: Seitz A, ed. Die Grossschmetterlinge der Erde. Stuttgart: Alfred Kernen, 457–
458.
Shirozu T. 1960. Butterflies of Formosa in colour. Osaka:
Hoikusha.
Shirozu T, Hara A. 1960. Early stages of Japanese butterflies
in colour. Vol. I-II. Osaka: Hoikusha.
Sidow A. 1992. Diversification of the Wnt gene family on
the ancestral lineage of vertebrates. Proceedings of the
National Academy of Sciences USA 89: 5098–5102.
Swofford DL. 1998. PAUP∗. Phylogenetic Analysis Using
Parsimony (∗and Other Methods). Version 4 ed. Sunderland,
Massachusetts: Sinauer Associates.
Teshirogi M. 1990. An illustrated book of the Japanese
Nymphalidae. Tokyo: Tokai University Press.
Wanntorp H-E, Brooks DR, Nilsson T, Nylin S, Ronquist
F, Stearns SC, Wedell N. 1990. Phylogenetic approaches
in ecology. Oikos 57: 119–132.
Weller SJ, Pashley DP, Martin JA. 1996. Reassessment
of butterfly family relationships using independent genes
and morphology. Annals of the Entomological Society of
America 89: 184–192.
Weller SJ, Pashley DP, Martin JA, Constable JL. 1994.
Phylogeny of noctuoid moths and the utility of combining
independent nuclear and mitochondrial genes. Systematic
Biology 43: 194–211.
Yagi N, Koyama N. 1963. The compound eye of Lepidoptera:
approach from organic evolution. Tokyo: Shinkyo-Press &
Co.
Zrzavy J. 1997. Phylogenetics and ecology: all characters
should be included in the cladistic analysis. Oikos 80:
186–192.
458
S. NYLIN ET AL.
APPENDIX 1: CHARACTER LIST
EXTERNAL MORPHOLOGY EXCEPT WING SHAPE AND
PATTERN
1. (a1) Shape of antennal club: (0) thick, short; (1)
intermediate; (2) thin, elongate. Ordered 012.
2. (a2) Colour of antennal club: (0) dark with bright
apical spot; (1) uniformly dark without bright apical
spot.
3. (ey1) Interfacetal hairs: (0) short and sparsely set;
(1) long and densely set.
4. (ey2) Radial connections between central and side
pupils: (0) absent; (1) weak; (2) strong. Ordered 012.
Data from Yagi & Koyama (1963).
5. (lp1) Apical segment of labial palp: (0) not distinctly
set off from subapical segment; (1) distinctly set off
from subapical segment.
6. (lp2) Colour of labial palp ventrally: (0) entirely
white; (1) with a black median stripe.
7. (lp3) Pubescence ventrally on labial palp: (0) hairs
absent; (1) uniformly dark, thin hairs; (2) dark, thick
bristles with whitish tip. Ordered 012.
8. (lp4) Dark hairs laterally on labial palp: (0) absent;
(1) present.
9. (le1) Colour of fore legs: (0) light; (1) light with dark
ventral band; (2) entirely dark. Ordered 012.
10. (le2) Position of secondary tooth of claws on hind
leg: (0) distinctly basad primary tooth; (1) laterad
or apicad primary tooth.
WING SHAPE AND PATTERN
See Figures 6–8 for position of some traits. Wing vein
terminology follows Niculescu (1965) and Scott (1986).
11. (ws1) Shape of median part of anterior margin of
hind wing: (0) straight or rounded, not incised; (1)
distinctly incised.
12. (ws2) Shape of posterior margin of fore wing: (0)
straight or convex, not incised; (1) slightly incised;
(2) strongly incised. Ordered 012.
13. (ws3) Process from outer margin of hind wing close
to M3: (0) absent; (1) small; (2) large. Ordered 012.
14. (ws4) Shape of outer margin of hind wing: (0) not
expanded posterior to M3; (1) distinctly expanded
posterior to M3.
15. (ws5) Shape of outer margin of fore wing: (0) not or
only slightly incised; (1) deeply incised.
16. (ws6) Process from outer margin of fore wing between M1 and M2: (0) absent; (1) present.
17. (ws7) Process from outer margin of fore wing close
to Cu2: (0) absent; (1) present.
18. (ws8) Shape of outer margin of fore wing: (0) more
or less even; (1) distinctly jagged.
19. (ws9) Hairs anteriorly on ventral side of fore wing:
(0) absent or only present basally; (1) extending
considerable distance from wing base.
20. (ws10) Vein R2 of fore wing: (0) issuing from cell
separately from R3+R4+R5; (1) separating from
R3+R4+R5 distal to cell.
21. (ws11) Wing span: (0) small, <50 mm; (1) large, >=
50 mm.
22. (ws12) Shape of anterior margin of hind wing: (0)
rounded distally, smoothly continuing in outer margin; (1) slightly incised distally; (2) strongly incised
distally. Ordered 012.
23. (ws13) Hairs basally on ventral side of hind wing:
Figure 6. Polygonia c-album: fore wing wing shape and
wing pattern characters. Arrows point to positions of
traits discussed in the text and numbers in parentheses
show the illustrated state of the character, conforming to
states in Appendices 1 & 2. Top half: dorsal side. Bottom
half: ventral side.
24.
25.
26.
27.
28.
(0) short and thin, hairlike; (1) long and thick,
bristlelike.
(wp1) Colour pattern on ventral side of hind wing:
(0) with contrasting bands and spots; (1) more uniformly coloured with part basad outer band of central symmetry system darker than rest of hind wing.
(wp2) Shape of outer band of central symmetry
system on ventral side of hind wing: (0) more regular,
running outside discal spot; (1) irregular, distinctly
incised medially, touching discal spot.
(wp3) Posterior part of lightly coloured discal spot
on ventral side of hind wing: (0) absent or not
distinguishable; (1) rounded or occasionally elongate, not angled; (2) elongate and angled, v-, u- or lshaped. Unordered.
(wp4) Ripple pattern on ventral side of fore and
hind wing: (0) absent; (1) present.
(wp5) Colour of wing veins basally on ventral side
PHYLOGENY OF NYMPHALINI BUTTERFLIES
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
Figure 7. Polygonia c-album: hind wing wing shape and
wing pattern characters. Arrows point to positions of
traits discussed in the text and numbers in parentheses
show the illustrated state of the character, conforming to
states in Appendices 1 & 2. Top half: dorsal side. Bottom
half: ventral side.
44.
45.
of hind wing: (0) dark; (1) white or yellow, contrasting with surrounding darker spots.
29. (wp6) (Subdivision of wp3:2) Shape of elongate and
angled posterior part of discal spot: (0) sharply
angled, v-shaped; (1) more roundedly angled, ushaped.
30. (wp7) Margin of discal spot: (0) indistinct, spot not
46.
47.
459
set off by narrow dark line; (1) distinct, spot at least
partly set off by a narrow dark line.
(wp8) Colour of discal spot: (0) white or yellow, not
metallic; (1) metallic silver or gold.
(wp9) Colour of ventral side of fore wing, between
M1 and M2: (0) light or intermediate, not contrasting
with surrounding areas; (1) dark, contrasting with
surrounding areas.
(wp10) Large white eyespots on ventral side of fore
wing: (0) absent; (1) present, at least the spot between M2 and M3.
(wp11) Colour pattern along basal part of anterior
margin of ventral fore wing: (0) uniform or with
more or less irregular pattern of dark and light
areas; (1) with regular, contrasting white and dark
lines.
(wp12) Colour pattern on ventral side of fore wing:
(0) different from ventral side of hind wing; (1)
similar to ventral side of hind wing.
(wp13) Shape of basal symmetry system on ventral
side of fore wing: (0) more or less evenly curved; (1)
weakly and roundedly bent medially; (2) strongly
and sharply bent medially. Ordered 012.
(wp14) Anterior part of basal symmetry system on
dorsal side of fore wing: (0) represented by continuous dark band or subcontiguous, square, dark
spots; (1) represented by well separated, rounded
spots.
(wp15) Colour pattern along anterior margin of fore
wing: (0) dark areas broken by at least one band of
lighter colour; (1) uniformly dark.
(wp16) Median eye spots on dorsal side of fore wing:
(0) present as contrasting white or red spots, at
least one of the spots between M1 and M3; (1) absent.
(wp17) Row of bright yellow-orange or blue, distinctly wedge-shaped spots immediately distad position of eyespots on dorsal side of fore wing: (0)
absent; (1) present.
(wp18) Row of contrasting bright yellow-orange or
blue spots or blue-green band immediately distad
position of eyespots on dorsal side of hind wing: (0)
absent; (1) present.
(wp19) Eye spots on ventral side of hind wing: (0)
small and simple or absent; (1) at least one eyespot
large, consisting of several concentric rings, but no
eyespot circular and completely closed; (2) at least
two eyespots circular and closed. Ordered 012.
(wp20) Blue band or series of blue spots immediately
outside parafocal elements on dorsal side of hind
wing: (0) absent; (1) present only posteriorly; (2)
percurrent. Unordered.
(wp21) Colour pattern on field along basal part of
anterior margin of dorsal fore wing: (0) uniformly
coloured or with weak stripes or mozaic pattern of
light and dark areas; (1) with distinct white and
black stripes.
(wp22) Colour of anterior spot or band immediately
distad central symmetry system on dorsal side of
fore wing: (0) white; (1) yellow to orange; (2) blue.
Unordered.
(wp23) Colour of anterior spot immediately distad
the g-system on dorsal side of fore wing: (0) white;
(1) yellow to red; (2) green. Unordered.
(wp24) Anterior part of discal spot on ventral hind
wing: (0) absent or indistinguishable; (1) dark, contrasting, well-defined elongate spot.
460
S. NYLIN ET AL.
Figure 8. Vanessa atalanta: fore wing and hind wing shape and pattern characters. Arrows point to positions of traits
discussed in the text and numbers in parentheses show the illustrated state of the character, conforming to states in
Appendices 1 & 2.
48. (wp25) (Subdivision of wp18) Colour of contrasting
spots or band immediately distad position of
eyespots on dorsal side of hind wing: (0) yelloworange; (1) blue or blue-green.
49. (wp26) Colour of eyespot or area around eyespot
between Cu1 and Cu2 on ventral side of hind wing:
(0) brown or orange, occasionally with some blue;
(1) yellow-green; (2) olive-green. Unordered.
50. (wp27) Colour of anterior part of central symmetry
system on ventral side of fore wing: (0) uniformly
dark or with large light and dark areas; (1) dark,
broken by narrow light lines of background colour;
(2) dark, broken by blue lines. Ordered 012.
51. (wp28) Colour of wing cell on dorsal fore wing: (0)
uniformly dark; (1) light, broken by vertical dark
bands or spots; (2) uniformly light, not broken by
dark spots. Ordered 012.
52. (wp29) Dark spot subapically between Cu2 and A2
on dorsal side of fore wing: (0) present; (1) absent.
53. (wp30) Dark band along posterior margin on ventral
side of fore wing, reaching or almost reaching outer
wing margin: (0) absent; (1) present.
54. (wp31) Colour of anterior, marginal dark spot corresponding to element g on dorsal side of fore wing:
(0) as dark as anterior, marginal dark spot corresponding to element f; (1) lighter than dark spot
corresponding to element f.
55. (wp32) Dark spot between Cu1 and Cu2 corresponding to element g on dorsal side of fore wing:
(0) present; (1) absent (or shifted out of this area).
56. (wp33). Seasonal polyphenism in wing colour:
(0) absent or not evident; (1) lightly coloured
(dominated by orange) dorsally in spring after
hibernation, dark (dominated by brown to black
with white band) in summer; (2) darkly coloured
(grey to black) ventrally in spring after hibernation, light (yellow to light brown) ventrally
in summer; (3) lightly coloured (orange) dorsally
in spring after hibernation, black areas present
at least on posterior part of hindwing in summer.
Unordered.
57. (wp34) Background colour outside central symmetry
system in posterior part of dorsal side of fore wing:
(0) blue; (1) white; (2) yellow to orange; (3) red; (4)
red-brown; (5) dark brown. Ordered 012345.
58. (wp35) Colour of anterior part of central symmetry
system on ventral side of hind wing: (0) dark; (1)
bright.
59. (wp36) Colour pattern along outer margin of dorsal
side of fore wing: (0) uniformly dark, or indistinct
darker and lighter areas; (1) distinct, alternating
white and black areas.
60. (wp37) Colour of dorsal side of hind wing basad
outer margin of central symmetry system, posterior
part: (0) bright or dark, not contrasting with areas
outside central symmetry system; (1) dark, contrasting with brighter areas outside central symmetry system.
61. (wp38) Colour just distad posterior part of central
symmetry system on dorsal side of fore wing: (0) not
PHYLOGENY OF NYMPHALINI BUTTERFLIES
contrasting yellow-orange to brown; (1) contrasting,
bright yellow.
INTERNAL MORPHOLOGY
Data and terminology mainly from Niculescu (1965).
62. (mg1) Shape of uncus: (0) with single point; (1) with
two points; (2) with many points. Unordered.
63. (mg2) Processus inferior and processus superior of
valva: (0) absent; (1) present.
64. (mg3) Fultura superior: (0) absent; (1) present.
65. (mg4) Shape of penis: (0) broad-based, distinctly
constricted distally, with narrow opening; (1) tubeshaped, not constricted distally, with wide opening.
Argynniti was coded as having the character nonapplicable because the penis shape differs markedly
from the two states found in the ingroup.
EGG MORPHOLOGY AND ECOLOGY
Data from general sources (see Material and Methods)
and our own investigations.
66. (e1) Number of vertical ribs on egg surface: (0) more
than ten; (1) ten or less.
67. (e2) Number of eggs in clutch: (0) one; (1) more than
two but less than twenty; (2) more than twenty.
Ordered 012.
68. (e3) Oviposition site: (0) leaves; (1) twigs.
69. (e4) Shape of egg clutch: (0) string; (1) heap; (2)
single layer. Unordered.
70. (e5) Colour of eggs: (0) yellow to brown; (1) green.
LARVAL MORPHOLOGY AND ECOLOGY
Data from general sources (see Material and Methods)
and our own investigations.
71. (L1) Larval nests: (0) none; (1) nest made out of leaf
or leaves; (2) communal web. Unordered.
72. (L2) Spiny clubs on larval head: (0) absent; (1)
present.
73. (L3) Middorsal spine on abdomen segment II-III:
(0) present (odd total number of spines per segment);
(1) absent (even total number).
74. (L4). Lightly coloured, contrasting, longitudinal
stripes on the larval abdomen between the middorsal and subdorsal spines. (0) present; (1) absent.
75. (L5). Lightly coloured contrasting pattern in the
shape of an inverted ‘V’ on the face of the larval
head. (0) absent; (1) present.
76. (L6). Large white patch or patches dorsally on posterior section of larval abdomen, contrasting with
orange anterior section. (0) absent; (1) present.
77. (L7). Alternating light and dark wedge-shaped spots
middorsally on larval abdomen: (0) absent; (1) present.
78. (L8) (Subdivision of L3:0) Colour of middorsal spine
anteriorly: (0) black; (1) yellow-orange; (2) greenish.
Unordered.
79. (L9) Colour of subdorsal spines: (0) black; (1) black
on thorax (t2) otherwise light; (2) light. Ordered
012.
80. (L10) Light, contrasting, lateral band on larva: (0)
present; (1) absent.
81. (L11) Secondary setae on the pinaculum on which L1
arises: (0) absent; (1) present. Data from Nakanishi
(1988).
82. (L12) Relative position of D1 and D2 on abdominal
83.
84.
85.
86.
87.
88.
89.
90.
461
segment 9: (0) D1 situated anteriorly to D2; (1) D2
situated anteriorly to D1. Data from Nakanishi
(1988).
(L13) Relative position of D2 and SD1 on abdominal
segment 9: (0) D2 dorsal to SD1; (1) SD1 dorsal to
D2. Data from Nakanishi (1988).
(L14) Secondary setae on abdominal segment 10, on
pinaculae from which P1 and Sp1 arises: (0) absent;
(1) present. Data from Nakanishi (1988).
(L15) Microspines on larval setae: (0) absent; (1)
present. Data from Nakanishi (1988).
(L16) Ornamentation of minute spines on larval
abdominal segment 10: (0) much; (1) restricted; (2)
absent. Ordered 012. Data from Nakanishi (1988).
(L17) Colour of silk spin: (0) white; (1) pinkish.
(L18) Position of filiform setae on abdominal segment 9 of larva: (0) arising from larval body surface;
(1) arising from the sclerotized base of the scolus.
Data from Harvey (1991).
(L19) Filiform setae on abdominal segments 1 and
2 of larva: (0) absent; (1) present. Data from Harvey
(1991).
L20. Subdorsal spines on abdominal segment 9 (penultimate segment) of last-instar larva: (0) two spines
present; (1) no spines present. Data from Teshirogi
(1990) and our own observations.
PUPAL MORPHOLOGY
Data from general sources (see Material and Methods)
and our own investigations.
91. (p1). Metal spots around base of dorsal spines and
in saddle of pupa: (0) present; (1) absent.
92. (p2). Dorsal projection on mesothorax of pupa: (0)
not or very slightly projecting; (1) raised into a sharp
point, (2) raised into a keeled projection. Unordered.
93. (p3). Subdorsal spines on abdominal segments of
pupa: (0) absent or very slightly projecting; (1) present, distinctly projecting.
94. (p4) Shape of pupal anterior projections: (0) very
slight projections; (1) projecting straight forwards
and to the sides, inner sides forming a straight line;
(2) projecting forwards, anterior section of inner
sides bending inwards, forming a curve. Ordered
012.
95. (p5) (Subdivision of p3:1) Length of subdorsal spines
on abdominal segment IV of pupa: (0) about as long
as spines on other segments; (1) distinctly longer.
ECOLOGY
96. (ec1). Hibernating developmental stage: (0) juvenile;
(1) adult. Data from general sources (see Material
and Methods).
97. (ec2). Host plant family: (0) Urticaceae; (1) Ulmaceae; (2) Cannabidaceae; (3) Salicaceae; (4)
Grossulariaceae (Ribes); (5) Betulaceae; (6) Ericaceae; (7) Asteraceae; (8) Malvaceae; (A) Acanthaceae; (B) Convolvulaceae; (C) Schrophulariaceae;
(D) Boraginaceae; (E) Verbenaceae; (F) Fabaceae;
(G) Rosaceae; (H) Violaceae.
Data from general sources (see Material and Methods)
and from a database compiled by N. Janz and S. Nylin
from the literature (see Janz & Nylin (1998) for sources)
and unpublished laboratory observations (Janz, Nyblom
& Nylin, 2001).
462
S. NYLIN ET AL.
APPENDIX 2: DATA MATRIX ON MORPHOLOGY, ECOLOGY AND BEHAVIOUR
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Argynniti
P. coenia
H. bolina
M. geoffroyi
Symbrentia sp.
A. levana
Hypanartia sp.
A. schaeneia
C. cardui
C. virginiensis
V. atalanta
B. gonerilla
I. io
A. urticae
N. milberti
N. antiopa
N. cyanomelas
N. polychloros
N. californica
N. xanthomelas
R. lalbum
K. canace
P. interrogationis
P. comma
P. progne
P. satyrus
P. gracilis
P. faunus
P. calbum
P. caureum
P. egea
1
2
3
A1 s A2 c ey1
4
ey2
5
Lp1
6
Lp2
7
Lp3
8
Lp4
9
10
11
Le 1 le2 s ws1
12
ws2
13
ws3
14
ws4
15
ws5
0
0
1
2
2
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
?
0
1
?
?
1
?
1
?
2
2
?
?
?
?
?
2
?
2
?
?
?
?
?
?
2
2
?
0
0
0
0
0
0
0
?
0
?
0
?
0
0
0
0
?
0
0
0
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
0
0
1
1
1
1
1
0
1
1
0
0
0
0
1
0
1
0
1
1
1
1
1
1
2
?
2
2
2
1
1
0
0
1
0
0
1
1
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
?
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
1
0
0
1
1
1
1
1
1
1
1
2
2
2
2
2
1
2
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
2
2
2
2
2
2
2
2
2
2
0
0
0
2
1/2
1
2
2
1
0
0
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
0
0
1
1
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
0
0
1
1
1
0
0
0
0
0
0
0
0
0
0
0
?
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
?
0
0
0
?
0
0
0
0
?
0
0
0
0
0
0
0
0
1
0
1
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
continued
463
PHYLOGENY OF NYMPHALINI BUTTERFLIES
APPENDIX 2 – continued
2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Argynniti
P. coenia
H. bolina
M. geoffroyi
Symbrentia sp.
A. levana
Hypanartia sp.
A. schaeneia
C. cardui
C. virginiensis
V. atalanta
B. gonerilla
I. io
A. urticae
N. milberti
N. antiopa
N. cyanomelas
N. polychloros
N. californica
N. xanthomelas
R. lalbum
K. canace
P. interrogationis
P. comma
P. progne
P. satyrus
P. gracilis
P. faunus
P. calbum
P. caureum
P. egea
16
ws6
17
ws7
18
ws8
19
ws9
20
21
22
23
24
25
26
ws10 ws11 ws12 ws13 wp1 wp2 wp3
27
28
29
30
wp4 wp5 wp6 wp7
0
0
0
0
0
0
0
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
0
1
1
1
1
1
1
1
1
0
1
0
0
0
0
0
0
0
0
0
0
1
1
1
1
?
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
1
0
?
0
?
0
?
0
0
0
0
?
0
?
0
0
0
0
0
?
?
?
?
0
?
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
0
0
1
1
1
1
1
1
1
0
0
1
1
1
1
1
1
1
1
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
0
0
0
0
0
1
0
1
0
0
0
1
1
1
1
1
0
1
1
1
2
2
2
2
2
2
2
2
2
2
2
0
0
0
0
1
1
0
1
1
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0
0
1
1
0
1
0
1
1
1
0
–
–
–
–
–
–
–
–
–
–
–
–
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
continued
464
S. NYLIN ET AL.
APPENDIX 2 – continued
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
wp8 wp9 wp10 wp11 wp12 wp13 wp14 wp15 wp16 wp17 wp18 wp19 wp20 wp21 wp22
3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Argynniti
P. coenia
H. bolina
M. geoffroyi
Symbrentia sp.
A. levana
Hypanartia sp.
A. schaeneia
C. cardui
C. virginiensis
V. atalanta
B. gonerilla
I. io
A. urticae
N. milberti
N. antiopa
N. cyanomelas
N. polychloros
N. californica
N. xanthomelas
R. lalbum
K. canace
P. interrogationis
P. comma
P. progne
P. satyrus
P. gracilis
P. faunus
P. calbum
P. caureum
P. egea
–
–
–
–
–
–
–
–
–
–
–
–
0
0
0
0
0
0
0
0
0
0
1
1
0
1
0
0
0
1
0
0
0
0
0
0
1
0
1
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
–
0
0
0
0
0
0
0
0
1
0
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
0
–
–
–
0
0
0
0
0
0
0
0
0
0
–
–
0
0
0
0
–
1
1
1
1
1
1
1
1
1
0
0
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0/1
0
1
1
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
1
0
1
1
2
2
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
1
1
1
1
0
2
2
0
2
2
1
2
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
–
–
1
1
0
0
1
0
0
0
1
1
0
–
1
1
1
1
2
1
1
1
1
1
1
1
1
1
continued
465
PHYLOGENY OF NYMPHALINI BUTTERFLIES
APPENDIX 2 – continued
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
wp23 wp24 wp25 wp26 wp27 wp28 wp29 wp30 wp31 wp32 wp33 wp34 wp35 wp36 wp37
4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Argynniti
P. coenia
H. bolina
M. geoffroyi
Symbrentia sp.
A. levana
Hypanartia sp.
A. schaeneia
C. cardui
C. virginiensis
V. atalanta
B. gonerilla
I. io
A. urticae
N. milberti
N. antiopa
N. cyanomelas
N. polychloros
N. californica
N. xanthomelas
R. lalbum
K. canace
P. interrogationis
P. comma
P. progne
P. satyrus
P. gracilis
P. faunus
P. calbum
P. caureum
P. egea
1
0
0
–
–
0
1
0
0
0
0
0
0
0
0
0
2
1
0
0
0
0
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
1
1
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
–
–
–
–
–
–
–
–
–
–
–
–
1
–
–
1
1
0
0
–
0
1
0
0
0
0
0
0
0
0
0
2
0
0
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
1
0
0
0
0
0
0
1
1
0
2
1
1
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
2
2
1
1
1
1
1
1
1
1
1
1
0
0
1
1
1
1
0
1
1
1
1
1
1
1
1
1
0
0
–
–
–
0
0
–
0
0
–
0
1
1
1
–
–
0
1
0
0
–
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
0
0
–
–
0
0
1
1
1
0
1
–
1
0/1
1
–
–
0
0
0
0
–
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
?
0
0
0
0
0
3
3
0
2
0
0
2
2
2
2
1
5
1
2
2
2
2
2
2
3
3
4
2
2
5
5
2
2
2
2
0
2
2
2
2
2
2
2
2
2
0
0
0
0
0
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
continued
466
S. NYLIN ET AL.
APPENDIX 2 – continued
61
62
63
64
65
66
wp38 mg1 mg2 mg3 mg4 e1 e
5
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Argynniti
P. coenia
H. bolina
M. geoffroyi
Symbrentia sp.
A. levana
Hypanartia sp.
A. schaeneia
C. cardui
C. virginiensis
V. atalanta
B. gonerilla
I. io
A. urticae
N. milberti
N. antiopa
N. cyanomelas
N. polychloros
N. californica
N. xanthomelas
R. lalbum
K. canace
P. interrogationis
P. comma
P. progne
P. satyrus
P. gracilis
P. faunus
P. calbum
P. caureum
P. egea
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
?
?
?
?
1
?
?
0
?
1
0
1
0
?
0
?
0
?
0
0
0
0
0
0
0
0
0
0
0
0
1
?
?
?
?
1
?
?
0
?
0
?
1
1
?
1
?
1
?
1
1
1
?
?
?
?
?
?
1
?
1
0
?
?
?
?
0
?
?
0
?
0
?
1
0
1
1
1
1
1
1
1
?
?
?
?
?
?
?
1
?
1
–
?
?
?
?
1
?
?
0
?
0
0
0
1
1
0
0
0
0
0
0
0
?
?
?
?
?
?
0
0
0
0
1
0
0
0
0
?
?
0
0
1
1
1
1
1
1
?
1
?
1
0
1
1
0
1
?
1
0
1
1
1
67
68
e2 n e3 e
0
0
0
2
2
2
0
0
0
0
0
0
2
2
2
2
?
2
2
2
1
0
1
1
0
0/1
0/1
0
0
0
0
–
0
0
0
0
0
0
?
0
0
0
0
0
0
0
1
?
1
0
1
1
0
0
0
0
0
0
0
0
0
0
69
70
e4 st e5 e
71
72
73
74
75
L1 la L2 s L3 m L4 tv L5 in
–
–
–
2
2
0
–
–
–
–
–
–
1
1
1
2
?
2
2
2
2
–
0
0
–
0
0
–
–
–
–
0
0
?
?
?
0
1
1
1
1
1
1
2
2
2
2
?
2
0
?
0
?
1
1
0
1
0
0
0
?
0
0
1
0
0
?
1
0
?
1
1
1
1
1
1
1
0
?
0
0
0
0
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
0
?
0
0
0
0
0
0
0
0
?
0
0
0
1
0
1
1
1
1
1
1
1
1
1
1
0
1
0
?
0
?
?
0
?
0
0
1
0
0
1
?
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
?
0
?
?
0
?
1
1
1
0
0
1
?
0
0
0
0
1
1
1
1
1
1
1
1
1
1
0
0
0
0
?
0
?
?
0
?
0
?
0
0
0
0
?
?
?
0
0
1
0
1
1
1
1
1
1
1
–
continued
467
PHYLOGENY OF NYMPHALINI BUTTERFLIES
APPENDIX 2 – continued
76
77
78
79
80
L6 p L7 ″ L8 c L9 s L10
6
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Argynniti
P. coenia
H. bolina
M. geoffroyi
Symbrentia sp.
A. levana
Hypanartia sp.
A. schaeneia
C. cardui
C. virginiensis
V. atalanta
B. gonerilla
I. io
A. urticae
N. milberti
N. antiopa
N. cyanomelas
N. polychloros
N. californica
N. xanthomelas
R. lalbum
K. canace
P. interrogationis
P. comma
P. progne
P. satyrus
P. gracilis
P. faunus
P. calbum
P. caureum
P. egea
0
0
0
0
?
0
0
?
0
0
0
0
0
0
0
0
?
0
0
0
0
0
0
0
0
0
0/1
1
1
0
0
0
0
0
0
?
0
?
?
0
?
0
0
0
0
0
0
?
0
0
0
0
1
1
1
1
1
1
1
1
1
1
–
0
–
0
?
0
?
?
1
?
0
2
–
0
0
0
?
1
1
–
0
1
1
1
1
2
1
1
1
1
1
2
0
2
0
?
0
0
?
2
?
0
2
0
0
0
0
?
2
?
0
0
2
1
2
1
2
2
2
2
2
2
0
0
0
1
?
0
?
?
0
0
0
0
1
0
0
1
?
0
0
1
0
0
0
0
0
0
0
0
0
0
0
81
L11
82
L12
83
L13
84
L14
85
L15
86
L16
87
L17
88
L18
89
L19
90
L20
?
1
?
?
0
0
?
?
0
?
0
?
0
0
?
0
?
?
?
0
1
0
1
?
1
1
1
?
1
1
?
?
0
?
?
1
1
?
?
1
?
1
?
0
0
?
0
?
?
?
0
?
0
1
?
?
1
?
?
1
1
?
?
0
?
?
0
0
?
?
0
?
0
?
0
1
?
1
?
?
?
1
0
0
0
?
?
0
?
?
0
0
?
?
0
?
?
1
1
?
?
0
?
0
?
0
0
?
0
?
?
?
0
0
0
0
?
?
0
?
?
0
0
?
?
1
?
?
1
1
?
?
0
?
0
?
1
0
?
1
?
?
?
1
0
0
0
?
?
0
?
?
0
0
?
?
0
?
?
1
1
?
?
2
?
2
?
2
2
?
2
?
?
?
2
0
0
0
?
?
0
?
?
0
0
?
?
0
?
?
?
0
?
?
0
?
0
?
0
0
0
0
?
1
?
?
?
?
1
1
1
1
1
1
0
?
0
0
1
1
?
?
?
0
?
0
0
0
0
0
0
0
0
?
0
0
?
?
0
0
0
0
0
0
0
0
0
0
0
0
0
?
?
?
1
?
1
1
1
1
1
1
1
1
?
1
1
?
?
1
1
1
1
1
1
1
1
1
1
0
0
0
1
1
1
–
–
0
0
0
0
0
0
0
0
–
0
0
0
0
0
0
0
0
0
0
0
0
0
0
continued
468
S. NYLIN ET AL.
APPENDIX 2 – continued
7
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Argynniti
P. coenia
H. bolina
M. geoffroyi
Symbrentia sp.
A. levana
Hypanartia sp.
A. schaeneia
C. cardui
C. virginiensis
V. atalanta
B. gonerilla
I. io
A. urticae
N. milberti
N. antiopa
N. cyanomelas
N. polychloros
N. californica
N. xanthomelas
R. lalbum
K. canace
P. interrogationis
P. comma
P. progne
P. satyrus
P. gracilis
P. faunus
P. calbum
P. caureum
P. egea
91
92
93
94
95
96
p1 m p2 d p3 a p4 p p5 p ec1
97
ec2
0
1
1
0
?
0
0
?
0
?
0
0
1
0/1
1
1
?
0
1
1
0
0
0
0
0
0
0
0
0
0
1
H
A&C&E
0&8&A&B
0
0
0
0&1
0
0&7&8&B&D&E&F
0&7&8&C&D&F
0
0
0&2
0
0
1&3&5&F&G
?
1&3&G
J
1&3
1&3&5
K
0&1&2
0&1&2
4&5&6
0&2&3
4&5&6
3&4&5&6
0&1&2&3&4&5
2
0
1
0
1
1
?
1
0
3
1
?
1
1
1
1
1
1
?
1
1
1
1
1
2
2
2
2
2
2
2
2
2
1
0
1
1
?
1
0
1
0
0
1
1
1
1
1
1
?
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
?
1
0
2
0
0
0
0
1
1
1
1
?
1
1
1
1
2
1
1
?
1
2
2
2
1
1
1
–
0
1
?
1
–
1
–
–
0
0
0
0
0
0/1
?
1
0/1
1
0
1
1
1
0
1
0
1
1
1
0
0
0
–
–
–
0
–
–
–
1
1
1
1
1
1
1
–
1
1
1
1
1
1
1
1
1
1
1
1
1
1